Intelligence at the Edge #8


Wireless home networking will be affordable, but will end users be buffaloed.

David's smiling face

By David S. Isenberg                                                               amnetlogo

From America's Network, April 1, 1999

I like Bluetooth — the wireless networking protocol designed to eradicate the hairball of cables behind your PC. It’s an Ericsson-IBM-Intel-Nokia-Toshiba scheme for cheap, unplug-and-play networks to link nearby devices. It’ll bring more intelligence to the edge.

Bluetooth will cost about $5 per end in a couple of years. It will be cost-competitive with physical cable, so you don’t necessarily need portability to benefit. It will be worth it just to eliminate the dust trap under your desk.

In the Bluetooth vision, your printer isn’t tied to your PC anymore; you can put it across the room if you want. Your modem becomes a clean, wireless package next to the RJ-11 jack. Your Palm Pilot syncs to your PC while it’s in your pocket. You connect your laptop to the net via a cell phone — while the phone is in your briefcase. You take your wireless Web-pad to the kitchen for coffee. You can do voice over Bluetooth, too, so your Bluetooth phone might deliver voice commands to your PC.

I can imagine other applications. Your phone flips Caller ID to your PC for a screen pop. Your car reports mileage, gas level and other stats when you come home. Your electric company reaches through your PC to your appliances for so-called demand-side energy management. Trivial apps, perhaps, but at five bucks an end, you can’t afford not to have them.

Despite these potential benefits, a lot of smart people bad-mouth Bluetooth:

• "It is oblivious to the IP [Internet Protocol] revolution," says ‘IP Everywhere’ evangelist Bob
• At 1 Mbps, Bluetooth is "much too slow to be useful," says wireless broadband expert Dewayne Hendricks of Com21.
• And Microsoft strategist Charles Fitzgerald voices concern with intellectual property and licensing. (Microsoft doesn’t even belong to Bluetooth’s 480-member special interest group.)

The Bluetooth protocol is, admittedly, a flawed idea. But then, so are single-family homes — horrendously energy-inefficient! So are cars, which cause 50,000 U.S. deaths per year. Then there’s television, Windows, the Republican Party, the Democratic Party, the telephone network and the modern corporation. Fortunately, flaws don’t keep stuff from happening.


HomeRF is another of the several wireless protocols vying for a place in the living room. It is a lot like Bluetooth — same 1 Mbps data rate, same 2.4 GHz band, same late-1999 product expectations and similar apps, with an equally powerful set of members, including Bluetooth’s IBM and Intel. Microsoft is a member, too.

Frankston likes HomeRF. He understands the power of emergent, unpredicted value. When he and Dan Bricklin wrote VisiCalc in 1978, the PC changed — by surprise, not design — from toy to tool. Frankston likes HomeRF because it incorporates native, end-to-end IP, the biggest creator of unpredictable value in our time. He thinks that Bluetooth is too proprietary and too preconceived to ramify as riotously as HomeRF could. He points to the overly constrained, often bungled, infrared communications standard, IRDA, as a cautionary tale.

I’m with Frankston on this. Open is better, layered is better, proven technology is better, end-to-end IP is better.


Putting my fridge online will be an easier decision if I can recoup the energy savings in a month or two. HomeRF will cost more than Bluetooth, but how much more is unclear. HomeRF is aimed at a PC-card implementation, while Bluetooth cleaves to an inherently cheaper embedded product model. Also, HomeRF is designed for a 50-meter transmission range. It needs a 100-milliwatt transmitter to do it. Bluetooth, in contrast, has a 10-meter range, and only needs one milliwatt. Dan Sweeney, general manager of Intel’s Home Networking Operations, says "there will always be a premium to pay for a HomeRF radio vs. Bluetooth."

Maybe Bluetooth’s nominal 10-meter range will prove overly limiting. Perhaps HomeRF’s other capabilities will become critical, such its support for five voice channels plus data, or its superior wall penetration, or end-to-end IP. I’m open.

Ultimately, easy-to-use protocols will win. I’m not going to become a networking guru so I can put my thermostat online.

"Because it is wireless, HomeRF sets high expectations for ease of use. You take it out of the box and turn it on," says Ben Manny, chair of the 65-member HomeRF Working Group. He could say the same for Bluetooth. Manny expects that both systems will benefit from the experience of wired home networking efforts now underway (e.g., HomePNA). If they’re cheap and stone-simple, I’ll like both. And when Hendricks’ Com21 gives us wireless home networks for digital video, I’ll like that too.

Copyright 1999 Advanstar Communications.